i8 7 6 



IIA\l)HO(IK OF PHYSIOLOGY 



M I !-'< IPIIYSH M l n.v. III 



capillary wall itself and the investing sheath of astro- 

 cytic cud feet. 



rschirgi (146) re-examined the two theories and 

 concluded that the perivascular glial membrane 

 was the more probable site of barrier activity. His 

 arguments were soundly belabored by Bakay (9), 

 Grontoft (60), Rodriguez-Peralta (127), Maynard 

 et al. (114, see above) and no doubt others who have 

 been unintentionally overlooked. Rodriguez-Peralta 

 (127) studied the histologic site of the hemato- 

 encephalic barrier with the aminoacridine dye 

 proflavine 1 1 C M which stains the nuclei of the cells 

 of all tissues except those of the central nervous system 

 after intravenous administration. The dye does not 

 enter the vascular endothelium of the cerebral blood 

 vessels, in contrast to the endothelial cells of other 

 tissues, and consequently he concludes that the 

 blood-brain barrier is localized in the endothelial cell 

 membrane facing the lumen. Since the dye penetrates 

 the choroidal epithelium but does not enter the 

 cerebrospinal fluid, he proposes that the cell mem- 

 brane of the choroidal epithelium facing the ven- 

 tricular lumen constitutes the blood-cerebrospinal 

 fluid barrier. Proflavine does penetrate the superficial 

 pia-glia membrane following subarachnoid adminis- 

 tration and also penetrates the intraneural vascular 

 endothelium. Because of this, and the prompt appear- 

 ance of the due in extraneural tissue following its 

 intrathecal administration, Rodriguez-Peralta sug- 

 gests that the neural vascular endothelium is 

 permeable from the direction of the central nervous 

 system, and the blood-brain barrier is thus polarized. 

 Intraventricular injection of the dye readily stained 

 the parenchyma adjacent to the ependymal walls, 

 indie, uint; that the ependymal cells do not impose a 

 significant barrier to this substance. 



Earlier investigations by Broman (20) using trypan 

 blue are in essential agreement with these results, 

 and there is widespread unanimity of opinion at the 

 presenl time that the essential barriei to the movement 

 of dyes into the central nervous system from the 

 plasma resides within the capillary endothelium 

 (1, 2, 9, 20, 6o, 1 ;;i 



Maynard et al. (114) present the following opinions 



regarding the blood-brain barrier, based on their 



electronmicrographs of the vascular bed of rat cerebral 

 cortex such as that here reproduced as figure 2. 

 "In seeking a locus for die blood-brain barrier, it 



in. iv be very significant thai the endothelial cells "I 



central nervous system capillaries form .1 completer) 



■ ontinuous layer without .mv suggestion of the fenes- 



ions thai have been seen in other capillaries such 



as arc well established in the kidney (123) and have 

 also been observed elsewhere. . . . Quite likely this is 

 a common feature of capillaries in the body gen- 

 erally, and the capillaries of the brain differ sig- 

 nilicantlyin having an uninterrupted endothelium. . . . 



"The present authors believe that the 'blood- 

 brain barrier' may be mainly an illusion. Physiologists 

 have been led to postulate its existence on the assump- 

 tion that there are considerable extracellular tissue 

 spaces which would be in equilibrium with blood 

 plasma, if a barrier did not exist. We see, however, 

 that the basic assumption is incorrect and that these 

 tissue spaces simply do not exist (135). 



"When the physiologist quantitatively studies 

 the penetration of ions, organic crystalloids, proteins, 

 vital dyes, etc. into nervous tissue, he must actually 

 be studying the penetration (or lack of it) into the 

 glial nervous cells. . . . Thus, we believe that the 

 physiological data available should be re-interpreted 

 on the assumption that nearly all substances are 

 within cytoplasmic compartments. L'nder these cir- 

 cumstances there would be no expectation of a simple 

 correlation with their distribution into the blood 

 plasma. The commonly accepted laws of permeability, 

 which apply elsewhere in the body, may thus indeed 

 apply in the brain without important exception and 

 without invoking a specialized barrier." 



Dempsey & Wislocki (30) on the basis of electron- 

 micrographs of rat brains following in vivo administra- 

 tion of silver nitrate, have implicated the capillary 

 basement membrane as the probable site of the blood- 

 brain barrier (figs. 2, 4). They observed silver deposits 

 within the endothelium of the capillaries and thought 

 that it could not proceed across the basement mem- 

 brane. Maynard et al. (1141 agree that the basement 



membrane is a continuous structure, but do not 

 believe that it differs from basement membranes found 

 in relation to capillaries elsewhere and are unwilling, 

 therefore, to emphasize its possible role in the blood- 

 brain barrier phenomenon. 



I less (72, 73) has described .1 homogeneous positive 

 periodic acid-Schifl" reaction between tin- cells, den- 

 drites, .ixon terminations and neuroglia libers of the 

 adult central nervous svsicm which he interprets as 

 indicating the presence of a carbohydrate-protein 

 ground substance. This positive periodic acid-Schiff 

 reaction appears in grev mattci of the mouse between 

 5 and 1 4 days postnatally, and is believed l>v I less to 

 determine the presence or absence of the ability of 

 intravenously administered trypan blue to stain the 



brain and, therefore, the absence or presence of the 



I ill Mid-brain barrier. 





